For binary JARs and Javadocs JARs https://sourceforge.net/projects/jann/files/jannlib-1.0.0-SNAPSHOT/ To view Javadocs http://jann.svn.sourceforge.net/viewvc/jann/jannlib/target/apidocs/index.html
Archive for category Linear Algebra
JANN First Release !
May 15
0.5435332331106713 0.1385960901595521 0.06904607041445548 0.0013574839878502721 2.5498285727219204E-4 7.382386611555016E-5 3.0226655138401433E-5 1.4509756496195933E-5 7.3475834110798174E-6 3.789949105647842E-6 1.9737860224435043E-6
This is a Comparison between Java Linear Algebra Libraries JAMA Apache Math library Jblas Performance Test (1): Functionality ( Multiplication ) M is 1000 *1000 N is 1000 * 1 Matrices Are Zeros !! ———————– Results: JAMA : 31 Milli Seconds Apache: 28 Milli Seconds Jblas: 5 Milli Seconds !! amazing right ? ———————– Performance Test [...]
====================== Headlines ====================== Hardlimit activation function Backpropagation FeedForward Network Six Layers Each Layer Contains 20 Neuron Input is 600 Node Elapsed time: 53 milliseconds , 2 Training Rounds RootMeanSquare=0.6250081520422739 END of NeuralNetworkSupervised Details. ====================== Snap shot on TestNetwork() Class ====================== –Weight=[[0.7664126978310184, 0.5967284054680337, 0.7087382499758025, 0.18991627253387666, 0.4157683454606339, 0.48961685375311526, 0.9640024042798034, 0.22160331733858796, 0.96028386987325, 0.3550337394097194, 0.08560567211451253, 0.02141483061456162, 0.29576195616793843, 0.27302235411023856, 0.2675171698936185, [...]
Here is the Matrix Multiplication Code import java.util.ArrayList; import java.util.Arrays; import java.util.Collection; import java.util.List; /** * @author marmoush * */ public class NNmatrix extends ArrayList<List<Double>> { /** * */ private static final long serialVersionUID = 3607185496007484139L; /** * */ public NNmatrix() { super(); } /** * * @param c */ public NNmatrix(Collection<? extends List<Double>> c) [...]
